autoevolution
 

Watch DeepRoute.ai Navigate Like a Seasoned Driver Through the Traffic Chaos of Shenzhen

DeeprRoute.ai navigates like a seasoned driver through the traffic chaos of Shenzhen 6 photos
Photo: DeepRoute.ai
DeeprRoute.ai navigates like a seasoned driver through the traffic chaos of ShenzhenDeeprRoute.ai navigates like a seasoned driver through the traffic chaos of ShenzhenDeeprRoute.ai navigates like a seasoned driver through the traffic chaos of ShenzhenDeeprRoute.ai navigates like a seasoned driver through the traffic chaos of ShenzhenDeeprRoute.ai navigates like a seasoned driver through the traffic chaos of Shenzhen
DeepRoute.ai has successfully demonstrated its Drive 2.0 software in downtown Shenzhen following Shenzhen’s Special Economic Zone decision to allow driverless cars on the road. The video shows DeepRoute.ai navigating challenging traffic scenarios without breaking a sweat.
Everybody might be talking about Tesla’s Full Self-Driving software and how GM or Ford might have better solutions. The hard truth is that all three companies might be left behind by the competition in China. A video shared on Youtube by DeepRoute.ai shows the impressive capabilities of their Drive 2.0 driverless software in complex traffic situations.

DeepRoute.ai has tested its Level 4 driverless system in Shenzhen, taking advantage of the city’s new regulations allowing driverless vehicles on the road starting August 1. Watching the software navigate complex traffic situations is impressive, considering the not-so-organized chaos. Two minutes into the video, we see bicycles, motorbikes, rickshaws, and cars coming from all directions at once, disregarding all traffic rules and safety concerns. And yet, DeepRoute.ai manages to find the perfect timing to cross through the intersection.

It’s taking its time, that’s for sure, with a 14-miles (22.5 km) route completed in one hour, but it manages to do so safely. Many situations encountered on the busy streets of Shenzhen would drive an average human driver to madness. Still, the autonomous driving system proves patient and forgiving.

According to DeepRoute.ai, the driverless vehicle is a modified MG Marvell R equipped with Drive 2.0, a production-ready L4 driverless software. The car leverages five solid-state LiDARs, eight cameras, and other sensors to scan the surroundings and achieve precise object detection up to 660 feet (200 meters). Its main advantage is that the software runs on a low-cost and power-efficient computing platform, which allows DeepRoute.ai to price it at an incredible $3,000 for automakers in mass production.

This is a significant price cut compared to where the company was in December 2021. At the time, they said the L4-capable Drive 2.0 would retail for roughly $10,000, which was still spectacular considering the hardware involved. Now, the company is working with carmakers to bring the first car models featuring the system as early as 2025. We’re sure the costs will get even more affordable by then.

If you liked the article, please follow us:  Google News icon Google News Youtube Instagram
About the author: Cristian Agatie
Cristian Agatie profile photo

After his childhood dream of becoming a "tractor operator" didn't pan out, Cristian turned to journalism, first in print and later moving to online media. His top interests are electric vehicles and new energy solutions.
Full profile

 

Would you like AUTOEVOLUTION to send you notifications?

You will only receive our top stories